Clinical profile of patients and sensitivity of troponin I in patients with and without acute coronary syndrome: An observational study

INTRODUCTION

Ischaemic heart disease remains a leading cause of death. Analysis of serum cardiac markers have become crucial for early diagnosis of acute coronary syndrome (ACS)¹ as electrocardiograph (ECG) can be non-diagnostic in 50%–60% of patients with MI coming to the emergency department and less than 80% have classical symptoms of chest pain. The most reliable of the cardiac marker assays, troponin assays, can be detected quickly and repeated serially with a half-life of 2 hours. These assays have become more sensitive, with the ability to quantify troponin levels even in the normal population, making the interpretation of this test difficult. Cardiac troponin (Tn) elevation reflects acute or chronic myocardial damage but is not exclusive for ACS, thereby causing difficulty in its interpretation.^2–4

There are various causes of Tn elevation in the absence of myocardial infarction, broadly due to oxygen demand mismatch (tachyarrythmias/bradyarrythmias, hypertensive crisis), direct myocardial damage (cardiac infiltrative disorders, chemotherapy) and systemic processes (sepsis, burns), etc.^3–6 It is important to exercise caution when interpreting elevated Tn values for diagnosing ACS in the absence of other features such as chest pain and ECG changes. Elevations in cardiac Tn, regardless of aetiology, portend a worse prognosis, even in those without heart failure (HF), obstructive coronary artery disease, or significant kidney dysfunction.⁷ The clinical significance of elevated troponins in non-cardiac conditions is not completely known. Also, it is unclear if these clinical conditions need to be managed differently when they lead to an increase in Tn levels.⁸

Identifying the cause for Tn elevation and its sensitivity and specificity in predicting ACS and cardiac mortality is an important first step in determining the optimal management for these patients. Hence, we proposed a hypothesis that among patients presenting with symptoms suggestive of ACS, TnI, as an index test has high sensitivity and specificity to diagnose ACS (target condition) and the study aimed to determine the sensitivity and specificity of TnI among medical inpatients diagnosed with ACS and to look for an alternative diagnosis in those with elevated TnI and no ACS.

METHODS

We conducted a retrospective study of adult inpatients who had TnI testing as part of their clinical assessment. The decision to test for TnI was at the discretion of the treating physician, for suspected ACS¹ and did not influence the inclusion of patients. We collected data from medical records and included adult inpatients who had TnI testing, either at admission or later, either at the emergency department (ED), or admitted in medical and cardiology wards, intensive care unit (ICU) or cardiac care unit (CCU). Patients were identified from the TnI test registers maintained in the respective departments of clinical biochemistry laboratory, CCU and ICU. In those patients who were diagnosed to have ACS, available data on echocardio-graphy/cardiac CT scan and the gold standard, coronary angiogram (CAG) were collected. Various departments were selected to have a heterogeneous population to minimize selection bias. Pregnant women were excluded. The diagnosis of ACS was made by the presence of two of the following: Symptoms of ischaemia and/or ischaemic changes on ECG and/or evidence of ischaemia on imaging, and elevated TnI.¹ The absence of these features, alongside the clinical picture of an alternative diagnosis, confirmed conditions other than ACS.

RESULTS

A total of 254 patients were included for analysis, 114 patients (45%) had normal TnI and 140 (55%) had elevated TnI. Twelve (10.5%) patients with normal TnI and 66 (47%) patients with elevated TnI had ACS. Of the 78 patients who had ACS, 66 (84.6%) had elevated TnI and 12 (15.38%) had normal TnI (Fig. 1).

Close

FIG 1.

Distribution of patients TnI troponin I ACS acute coronary syndrome

Export to PPT

The mean (SD) age of all patients was 56.67 (16.87) years. The mean (SD) age of patients with ACS irrespective of TnI was 62.9 (12.0) years and those without ACS was 54 (18) years (p<0.001). Of the 254 patients, 156 (61.4%) were men and 98 (38.6%) were women. The baseline characteristics of patients across the four groups showed significant differences. Age, history of chest pain and smoking, and ICU admission were higher in patients with ACS with or without elevated TnI. Comorbid conditions were higher in patients with elevated TnI, with or without ACS. Of the 254 patients, 106 (41.73%) had diabetes and mean (SD) duration of diabetes was 11.7 (8.6) years. One hundred and twenty-four patients (48.8%) had hypertension and the mean (SD) duration of hypertension was 10 (8.3) years, 58 (22.8%) were known to have ischaemic heart disease, of which 27 (10.6%) had undergone coronary revascularization (Table I).

DISCUSSION

Our study details the clinical profile and causes of elevated TnI and outcomes of 254 inpatients from a tertiary care teaching hospital. Of the 254 patients, more than half (140, 55%) had an elevated TnI level. And among these 140 patients with elevated TnI levels, half (74, 53%) had diagnoses other than ACS, a higher number in comparison with an audit done by Wong et al.¹¹ Male predominance in the ACS group was noted in this study as also seen in the other studies.^3,11,12 No significant difference was noted between the gender distribution of patients with normal and elevated troponin. We had 47% of patients with elevated TnI having ACS, lower compared to the study by Wong et al. (62.3%).¹¹ We speculate that this difference in the finding between our study and other similar studies could be due to a more sensitive assay used by us leading to detection of minor elevations of troponin even in non-ACS patients with an alternative diagnosis for TnI elevation. There were no significant differences with the medical history and comorbid conditions in the two groups (normal v. elevated TnI).

In our study, patients with ACS and elevated TnI had a significant association with diabetes (53.2%) and history of smoking (70.5%) compared to the study by Khan et al.,³ probably due to lower numbers in their study. Among patients with elevated TnI, we found that more patients with ACS had pre-existing hypertension (53.8%), ischaemic heart disease (51.4%) and chest pain (81.8%) and presence of ST elevation on ECG at presentation than without ACS, similar findings were seen in another study (54%, 45% and 72%, respectively).³ Patients without ACS were less likely than those with ACS to present with ECG changes and chest pain as a major symptom (p<0.001), as noted in a similar study.¹³

A large number of conditions have been identified for elevation of TnI, other than ACS. Among the 140 patients (55%) with elevated TnI in the absence of ACS, sepsis, AKI and HF were the commonest causes in our study. Significantly higher number of patients and with similar non-ACS aetiologies of elevated TnI were seen in other studies.^14,15 The less common causes of elevated TnI in the absence of ACS were myocarditis, cardiac arrest from non-ACS causes, chronic kidney disease, accelerated hypertension and cardiotoxic drugs. We had one patient with pulmonary thromboembolism and 4 patients in whom no diagnosis could be made.

The mean duration of hospital stay in patients with elevated TnI with or without ACS was significantly higher than in those with normal TnI (8.45 [5.9], 5.81 [3.5] days). In-hospital all-cause mortality and cardiac mortality were significantly higher in patients with elevated TnI and ACS (23% and 13.6%) than elevated TnI without ACS (12% and 4%), respectively. This is in contrast to other studies in which non-ACS patients with elevated Tn had significantly higher mortality.^11,14 This could be due to the larger sample sizes in these studies.

The sensitivity and specificity of TnI for the diagnosis of ACS was 84.6% and 58%, respectively in our study. These numbers are low compared to other studies as they had excluded non-ACS conditions.^15,16 The result may be confounded by the presence of non-ACS aetiologies in patients with ACS in this study. Elevated TnI is an independent predictor of prognosis in patients with ACS.^17,18 It is also associated with higher mortality in patients with non-ACS aetiologies.^19–22 We found the sensitivity and negative predictive value (NPV) of TnI in predicting all-cause and cardiac mortality to be (96% and 100%) and (99.12% and 100%), respectively.

A few limitations of the study are small sample size, qualitative assessment of TnI, which limit the assessment of possible incremental mortality with increase in TnI and the limited data on coronary angiogram.

In conclusion, TnI can be elevated in several conditions apart from ACS. The assessment of whether TnI elevation is a result of a coronary event or not has become a major challenge and has greater implication to both the hospital and the patient. Caution should be exercised while diagnosing a patient with elevated TnI as having ACS without other supporting evidence and in the setting of an alternative diagnosis. Specificity of TnI can be improved by better test-ordering protocols based on the pre-test probability of the suspected condition. However, TnI has a good negative predictive value in diagnosing ACS and has a role in predicting mortality in patients with and without ACS. Elevated TnI portends a worse prognosis regardless of the aetiology, compelling for a detailed evaluation of both ACS and non-ACS causes. As elevation of TnI indicates myocardial injury even in those with non-ACS cause of its elevation, whether they require longer follow-up and early cardiac evaluation later in life has to be studied.